So that’s not precisely true, I mean it’s precise historically speaking but you have slightly more options nowadays with modern browsers. I’m not saying it’s good or easy options but there are.

One option would be to download the files to the browser’s memory with say JavaScript fetch API Using the Fetch API - Web APIs | MDN . Once you have the file in memory you can then save the file to the filesystem with file system JavaScript API (example: https://jsfiddle.net/koldev/cW7W5/).

One could approach this by concatenating the files server side, with some boundary marker in between, then download the files to browser memory, split it by boundary marker into multiple Blobs, then for each Blob trigger saving to filesystem.

I think at this point browsers (at least Chrome) would actually warn user that “this web site is trying to download multiple files” within a top bar pop up but user can still allow it.

Overall, I do not think it’s worth it. Just do multiple HTTP calls from web site to the back-end, or use ZIP to package multiple files into one.